Automated subcortical segmentation using FIRST: Test-retest reliability, interscanner reliability, and comparison to manual segmentation

Abstract

Multiple techniques exist for the automated segmentation of magnetic resonance images (MRIs). The validity of these techniques can be assessed by evaluating test-retest reliability, inter-scanner reliability, and consistency with manual segmentation. We evaluate these measures for the FSL/FIRST subcortical segmentation tool. We retrospectively analyzed 190 MRI scans from 87 subjects with mood or anxiety disorders and healthy volunteers scanned multiple times on different platforms (N=56) and/or the same platform (N=45, groups overlap), and 146 scans from subjects who underwent both high-resolution and whole brain imaging in a single session, for comparison with manual segmentation of the hippocampus. The thalamus, caudate, putamen, hippocampus, and pallidum were reliably segmented in different sessions on the same scanner (ICC>0.83 scanners and diagnostic groups pooled). In these regions, the range of between-platform reliabilities were lower (0.527<ICC<0.953), although values below 0.7 were due to systematic differences between platforms, or low reliability in the hippocampus between eight- and single-channel coil platforms. Accumbens and amygdala segmentations were generally unreliable within and between scanning platforms. ICC values for hippocampal volumes between automated and manual segmentations were acceptable (ICC>0.7, groups pooled), and both methods detected significant differences between genders. Additionally, FIRST segmentations were consistent with manual segmentations (in a subset of images; N=20) in the left caudate and bilateral putamen. This retrospective analysis assesses realistic performance of the algorithm in conditions like those found in multi-site trials or meta-analyses. Additionally, the inclusion of psychiatric patients establishes reliability in subjects exhibiting volumetric abnormalities, validating patient studies.

Hippocampal Atrophy in Major Depression: a Function of Childhood Maltreatment Rather than Diagnosis?

“…Also, the results of our study were demonstrated by the use of two independent, complementary methods of subcortical morphometry (FIRST, VBM8). FSL-FIRST was shown to provide precise segmentations with high reliability for the hippocampus and has successfully been utilized in MDD studies (Lai and Wu, 2011;Nugent et al, 2012), and using FSL-FIRST allowed displaying not only overall volume information, but also localized differences in shape. Furthermore, we aimed to obtain gray matter information of the hippocampus by the use of VBM, as it represents one of most commonly used instruments in morphometric studies and is highly sensitive for detecting morphometric changes in depression (Bergouignan et al, 2009) and maltreatment alike (Chaney et al, 2014;Dannlowski et al, 2012).…”

“…fmrib.ox.ac.uk/fsl/fsl-4.1.9/first/index.html). Surface meshes representing the volumetric information of each structure are formed using parameterized deformable surfaces (Nugent et al, 2012). FIRST then ascertains the most probable shape by means of linear combinations of shape variations employing the learned models.…”

Association of Depressive Symptoms with Hippocampal Volume in 1936 Adults

“…In addition, the study controlled for medication use, alcohol use, education, and other variables that might influence the relationship between depression and hippocampal volume. Another potential strength is the use of an automated method to derive hippocampal volumes that results in excellent reproducibility (Lucarelli et al, 2012;Nugent et al, 2012) and avoids a left-right bias that may be inherent with manual segmentation (Maltbie et al, 2012).…”

Robust whole-brain segmentation: Application to traumatic brain injury

“…The assumption is that brains of healthy subjects do not change substantially within short periods. ICC is widely used to quantify test-retest reliability (Kempton et al, 2011;Nugent et al, 2013).…”